AN EXPERIMENTAL TEST OF A THEORY OF RESPONDENT BEHAVIOR.
Abstract
An experiment was conducted in which methodologically imperfect conditions conducive to content-irrelevant responding were created. Subjects' responses to unstructured questionnaires given under 'normal' test conditions were compared with responses obtained from the same subjects given to structured questionnaires under 'biased' conditions. Judges were used to determine the degree or extent of deviation in responses given to the subsequent or experimental questionnaires. The results indicate first that the use of structured questionnaires increases subjects' evaluation apprehension in that subjects were found to endorse socially-desirable statements more than undesirable statements. Second, subjects, in general, yielded to the biasing effects of the social aspects of the research situation and endorsed both socially-desirable and undesirable items, deviating from their original positions. Third, interest in item-content reduced subjects' content-irrelevant responding, even under explicit social pressures. Fourth, the effects of instrument and personality variables were, to a large degree, dependent on other situational factors, such as interest in itemcontent and the relevancy of the social aspects of the research situation. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1965
- Accession Number
- AD0622504
Entities
People
- Engin Inel Holmstrom
Organizations
- Duke University